Arrhythmias & electrophysiology
External validation of a machine learning-based classification algorithm for ambulatory heart rhythm diagnostics in pericardioversion atrial fibrillation patients using smartphone photoplethysmography: the SMARTBEATS-ALGO study
Authors' conclusion
A machine learning-based algorithm demonstrated excellent performance in diagnosing atrial fibrillation and atrial flutter from smartphone-PPG recordings in an unsupervised ambulatory setting, minimizing the need for manual review and ECG verification, in elderly cardioversion populations.
Study context
- Population
- Patients undergoing cardioversion for atrial fibrillation (AF) or atrial flutter (AFL), recording heart rhythm in an unsupervised ambulatory setting280 in external validation cohort
- Participants
- 460 patients
- Comparison
- Smartphone-PPG machine learning (ML) algorithm vs simultaneous single-lead electrocardiogram (ECG)
- Outcome
- Diagnostic performance (sensitivity, specificity, accuracy) for AF and AF/AFL vs simultaneous ECG
Methodological detailsexternal validation, unsupervised ambulatory setting
- External validation
- Unsupervised ambulatory setting
Performance
As reported in the abstract
Diagnostic performance vs simultaneous single-lead ECG (gold standard), external validation cohort
| Test or model | AF sensitivity | AF specificity | AF accuracy | AF/AFL sensitivity | AF/AFL specificity | AF/AFL accuracy |
|---|---|---|---|---|---|---|
| PPG | 99.7% | 99.7% | 99.7% | 99.3% | 99.1% | 99.2% |
PPG
AF sensitivity99.7%
AF specificity99.7%
AF accuracy99.7%
AF/AFL sensitivity99.3%
AF/AFL specificity99.1%
AF/AFL accuracy99.2%
Original publication
External validation of a machine learning-based classification algorithm for ambulatory heart rhythm diagnostics in pericardioversion atrial fibrillation patients using smartphone photoplethysmography: the SMARTBEATS-ALGO study. Europace. Published 17 February 2025. doi:10.
Summarised from the published abstract; every number is checked against it